Various processes have been devised for the measurement of the chemical components of a complex mixture separated by gas, liquid or supercritical fluid carriers, and for the measurement of the composition of gas, liquid and supercritical fluid streams or the gases evolved upon heating a solid matrix. Representative of such processes are detection by changes in physical properties of the streams, including changes in the refractive index and thermal conductivity. Another detection scheme is based on the measurement of electrical currents induced by the formation of ionic species during combustion of a stream (flame ionization). Irradiation of a stream using electromagnetic radiation and radioactive sources or changes in the absorption of electromagnetic radiation by components of a stream are also employed.
In general, these processes allow detectors that can be classified as either "general" or "universal" detectors. These detectors produce a response for all of the chemical constituents contained in a carrier stream (except for the eluent or carrier itself). Selective detectors, on the other hand, respond to specific chemical constituents based on one or more elements within each compound detected and/or unique physical or chemical properties of the components. Selective detection is often required when the chemical components of interest are present at low concentrations, together with much higher concentrations of other chemicals in the stream.
Detection systems can be further classified as being non-destructive detectors, in which the chemical composition of the stream is not altered by the measurement process, or destructive detectors, in which the sample is destroyed or chemically altered as a result of the measurement process. Generally, to use a destructive detector, such as a flame ionization detector, in combination with a selective detector, it is necessary to split the sample stream prior to measurement of the chemical constituents by the respective detectors. Difficulties in controlling the amount of the sample stream which flows into the different detectors using stream splitting has severely limited the utility of this technique.
In chromatographic analysis, the identity of a chemical compound is determined based on the "retention time" of the compound in a chromatographic system. The amount of the compound is determined based on the detector response. Typically, several analyses are performed using standard solutions of the test compound at different concentrations. Based on this information, a calibration curve is constructed by comparing the detector response (e.g. peak area) to the amount of injected compound. For both "universal" and "selective" detectors, retention time as well as response of a given chemical compound relative to a "standard" compound provide information regarding the identity of the chemical.
Once response factors for a wide range of chemical compounds are known, it is possible to determine the concentrations of different compounds based solely on their retention times and detection response without the need for constructing calibration curves for each individual component. For example, relative response factors using a flame ionization detector are available for a large number of hydrocarbons and other chemical compounds found in petroleum and petrochemical samples, thus greatly simplifying quantifications of these complex samples. Comparison of the relative response factors for compounds on two or more different detectors provides even more information regarding the identity of a particular chemical compound, since fundamentally different measurement techniques are employed.
An important class of selective detectors are devices for the selective measurement of sulfur-containing compounds. When present as impurities at low concentrations, sulfur-containing compounds are detrimental to a wide range of chemical processes. In consumer products, trace levels of sulfur-containing compounds can impart objectionable taste and odor to products. In petrochemical applications, trace sulfur contaminants can rapidly poison costly catalysts. For these reasons, numerous processes and apparatus have been developed for the measurement of low concentrations of sulfur-containing compounds in sample streams.
Representative of such processes is that disclosed in West German Patent No. 1,133,918 to H. Dragerwerk and B. Drager for the flame photometric detector (FPD). Sulfur-containing compounds are oxidized in a hydrogen/air flame to form diatomic sulfur S.sub.2 in an electronically excited state. Emission of light from this species can be measured using a photomultiplier tube or similar light detection device equipped with an optical filter to eliminate the light emitted from other species in the flame. Through the use of different optical filters, the FPD can also be used for the measurement of phosphorus-containing compounds based on the emission of light from electronically excited phosphorus dioxide (PO.sub.2) formed in the hydrogen/air flame. Compounds that do not contain sulfur or phosphorus cannot be measured using the FPD. However, compounds that do not contain sulfur or phosphorus can result in a decrease or "quenching" of the detector response for sulfur-and phosphorus-containing compounds. S. O. Farwell and C. J. Barinaga, Sulfur-Selective Detection with the FPD: Current Enigmas, Practical Usage, and Future Directions, 24 Journal of Chromatographic Science, 483 (1986).
The reactive sulfur and phosphorus species generated in the FPD flame are short-lived and therefore require that the light measurement device be located in close proximity to the flame. This basic design requirement has precluded the simultaneous operation of the FPD with other "universal" detection systems, such as the flame ionization detector. Of course, there is no such thing as a truly universal detector. For example, the flame ionization detector responds sensitively to nearly all organic compounds (excluding formaldehyde and formic acid) but not to inorganic compounds (e.g., O.sub.2, N.sub.2, Ar, CO.sub.2, CO, SO.sub.2, H.sub.2 S, COS, etc.).
Another example of selective detection of sulfur-containing compounds is that disclosed in U.S. Pat. Nos. 4,678,756 and 4,352,779 of R. E. Parks. According to this process, a sample is passed through a furnace containing a metal oxide catalyst to convert sulfur-containing compounds to sulfur dioxide. The sulfur dioxide is then passed through a second furnace, where the sample is mixed with hydrogen gas to facilitate the conversion of sulfur dioxide to hydrogen sulfide. The effluent of the second furnace is then directed into a reaction cell where the hydrogen sulfide is mixed with ozone and the resultant chemiluminescence is measured by means of a photomultiplier tube. In the system described by Parks, non-sulfur-containing compounds cannot be measured, since the expected products from the oxidation furnace (carbon dioxide and/or carbon monoxide) and from the reduction furnace (methane) either do not undergo an ozone-induced chemiluminescent reaction, or the light emitted from such reactions is eliminated through the use of optical filters.
Numerous other patents and publications may be found which disclose other approaches to the selective detection of sulfur-containing compounds. Gaffney and co-workers have described a technique for the measurement of reduced sulfur compounds (e.g., hydrogen sulfide, methanethiol, dimethyl sulfide, etc.) based on reactions of the sulfur-containing compounds with ozone to form electronically excited sulfur dioxide (SO.sub.2 *) which then emits radiation in the 200 nm to 400 nm region of the spectrum. J. S. Gaffney, D. F. Spandau, T. J. Kelly, R. L. Tauner, Gas Chromatographic Detection of Reduced Sulfur Compounds Using Ozone Chemiluminescence, 347 Journal of Chromatography 121 (1985). This detection system does not permit measurement of all sulfur-containing compounds (e.g., sulfur dioxide) and suffers from interferences from non-sulfur-containing compounds such as olefins.
Birks and co-workers have described a sulfur selective detector based on fluorine-induced chemiluminescence. J. K. Nelson, R. H. Getty, J. W. Birks, Flourine Induced Chemiluminescence Detector for Reduced Sulfur Compounds, 55 Analytical Chemistry 1767 (1983). Reduced organic sulfur-containing compounds (e.g., mercaptans, sulfides, disulfides, etc.) react with molecular fluorine to form vibrationally excited hydrogen fluoride and other electronically and vibrationally excited species which emit radiation in the red and near infrared region of the spectrum. Inorganic sulfur-containing compounds (e.g., H.sub.2 S, SO.sub.2, etc.) do not undergo fluorine-induced chemiluminescence, while many non-sulfur-containing compounds, such as olefins and aromatic hydrocarbons, do react and interfere in the measurement of sulfur compounds.
Other workers have described a detection system based on the reaction of sulfur-containing compounds with chlorine dioxide. These reactions result in the formation of electronically excited diatomic sulfur, which emits radiation in the visible region of the spectrum (250 to 450 nm).
None of these previously reported systems for the measurement of sulfur-containing compounds permit the simultaneous measurement of non-sulfur-containing compounds, without the need for splitting the sample to a second "universal" detector system.
Halstead and Thrush have described the chemiluminescent reaction of sulfur monoxide with ozone. The Kinetics of Elementary Reactions Involving the Oxides of Sulphur III. The Chemiluminescent Reaction Between Sulphur Monoxide and Ozone, C. J. Halstead, B. A. Thrush, 295 Proceedings of the Royal Society, London, 380 (1966). Sulfur monoxide, produced from sulfur dioxide using a microwave discharge, was reacted with ozone. One of the reaction products was identified as electronically excited sulfur dioxide. The emission spectrum from this species was recorded and found to extend from 280 to 420 nm, with maximum emission at 350 nm.
Previous studies have shown that sulfur monoxide is one of the species formed in the combustion of sulfur compounds in a flame. Sulfur Chemistry in Flames, C. H. Muller, et al., in 17th International Combustion Symposium, pp 867-879 (1989) and Experimental and Numerical Studies of Sulfur Chemistry in H.sub.2 /O.sub.2 /SO.sub.2 Flames, M. R. Zachariah, O. I. Smith 69, Combustion and Flame 125 (1987). Under the typical operating conditions of the FPD, sulfur monoxide is present at about 10 times the level of diatomic sulfur.
On the basis of these observations, Benner and Stedman reported the development of a "Universal Sulfur Detector" (USD) based on the formation of sulfur monoxide in a hydrogen/air flame and subsequent detection of SO based on a chemiluminescent reaction with ozone. R. L. Benner, D. H. Stedman, Universal Sulfer Detection by Chemiluminescence, 60 Analytical Chemistry 1268 (1989). The original embodiment of the USD was a continuous monitor for the measurement of the total concentration of sulfur-containing compounds in ambient air. The USD is also described in the parent U.S. patent application Ser. No. 07/275,980. In this design, the air stream containing the sulfur compounds is mixed with an excess of hydrogen in a quartz burner assembly equipped with an external ignition source. A quartz sampling probe is used to collect sulfur monoxide and other products from the flame for transfer to a modified nitric oxide/ozone chemiluminescence detector.
This detection system was found to provide greater sensitivity for the measurement of sulfur-containing compounds than existing sulfur-selective detectors and did not suffer interferences in the measurement of sulfur species due to the presence of higher concentrations of non-sulfur species such as water, carbon dioxide and heptene.
The USD is designed so that the sulfur-containing compounds are contained in an air stream which is required to support combustion when mixed with a second stream containing hydrogen gas. The optimum gas flow rates were determined to be 400 to 500 mL/min of air and 300 mL/min of hydrogen. The optimum internal diameter for the quartz sampling probe was reported as about 0.1 mm. A key feature of the system reported by Benner and Stedman was the need to add a halogenated compound, such as CF.sub.2 Cl.sub.2, into the air stream in order to achieve stable instrument baseline and long term instrument stability. The fundamental design of the USD precludes the measurement of non-sulfur containing compounds by conventional means such as the detection of ionic species produced in the flame.